Palmlike Plants from the Dolores Formation (Triassic) Southwestern Colorado

GEOLOGICAL SURVEY PROFESSIONAL PAPER 274-H Palmlike Plants from the Dolores Formation (Triassic) Southwestern Colorado

By ROLAND W. BROWN

A SHORTER CONTRIBUTION TO GENERAL GEOLOGY

GEOLOGICAL SURVEY PROFESSIONAL PAPER 274-H

UNITED STATES GOVERNMENT PRINTING OFFICE, WASHINGTON : 1956 UNITED STATES DEPARTMENT OF THE INTERIOR Douglas McKay, Secretary

GEOLOGICAL SURVEY Thomas B. Nolan, Director

For sale by the Superintendent of Documents, U. S. Government Printing Office Washington 25, D. C. CONTENTS

Page Page Abstract.-______205 Relationships of Sanmiguelia^-- ______207 Geologic occurrence of the plants.______205 Bibliography.______-_-___-___-_-----_- 209 Description of the specimens,______206

ILLUSTRATIONS

Page FIGURE 29. Reconstruction of Sanmiguelia lewisi Brown______206 PLATES 32, 33. Triassic plants from southwestern Colorado ______------_-- _ -Following 206 in

A SHORTER CONTRIBUTION TO GENERAL GEOLOGY

PALMLIKE PLANTS FROM THE DOLORES FORMATION (TRIASSIC) IN SOUTHWESTERN COLORADO

By KOLAND W. BROWN

ABSTRACT Partial section at old Fall Creek post office, Colorado Plant remains from the Dolores formation in southwestern Colorado include a few small twigs of the conifer Brachyphyllum Jurassic and large, many-ribbed leaves described as those of an angio- Entrada sandstone. 50 ft. Whitish massive sandstones with sperm, tentatively regarded as a primitive palm and named few dark, shaly or limy thin-bedded strata. Sanmiguelia lewisi. This is the earliest known angiospermous Triassic flowering plant. Its occurrence in the Middle to Late Triassic, Dolores formation. 485 ft. Bright red, calcareous, massive however, indicates that evidence for the origin of the angio- and thin-bedded cliff-forming sandstones, few shales, sperms should be looked for in the earlier Triassic or late and some conglomerates. It yields remains of meto- Paleozoic. posaurs, phytosaurs, and plants. The lowermost whitish conglomeratic sandstone rests on an erosion GEOLOGIC OCCURRENCE OF THE PLANTS surface of the Cutler formation. Permian In September 1953, G. Edward Lewis, of the U. S. Cutler formation. 275 ft. Dark red shales and sandstones, Geological Survey, searching for fossil remains in the with fragmentary plants. upper part of the San Miguel Kiver valley in south­ The Dolores formation, named and described by western Colorado, found a palmlike leaf impression in Whitman Cross (1899) from exposures in the valley of the red Dolores formation near Placerville. This Dolores River in southwestern Colorado, at first specimen, 38 cm long and 20 cm wide, was fragmentary, included the red beds that are now known as the Cutler lacking the apex and base. Therefore, further search formation. However, the discovery in 1904 near was made at the same locality early in 1954, but Lewis Ouray of an angular unconformity below the fossili- found only a few more fragments, none of which gave ferous horizon in the Dolores caused Cross and his a clue to the nature of the basal or petiolar portion of associates (Cross, Howe, Ransome, 1905; Cross, the leaves. However, in September 1954, Lewis and Ransome, 1905; Cross, Howe, Irving, 1907) to restrict the writer, at an outcrop one-half mile north of the the name Dolores to the Triassic strata and to name original site, succeeded in finding fairly good specimens the more reddish underlying beds the Cutler formation, that illustrate the salient features of these leaves, provisionally of Permian age. Cross reported that some of which were still attached to part of a structure­ R. C. Hills in 1880 and 1882 had announced the dis­ less cast of a stem. Further exploration of the fossil- covery of fossils in the Dolores, including reptilian bearing bed along its outcrop for 10 miles on both sides teeth, ganoid fishes, mollusks, ana plants. Before of the San Miguel Kiver yielded equally good material these could be described, however, they were lost. to the writer in June, 1955. Cross said that in the course of his survey of the area The principal locality is in SE% sec. 12, T. 43 N., R. more reptilian teeth and a plant, identified by David 11 W., in a low cliff of reddish calcareous sandstone, on White as Pachyphyllum [an error for Brachyphyllum] the hillside of the right bank, 420 feet above the San munsteri, were found. All of these fossils reputedly Miguel Kiver and 145 feet above the base of the indicated the Triassic age of the Dolores. At the Fall Dolores formation, near the old Fall Creek post office, Creek post office locality in the Dolores the writer and about 2% miles southeast of Placerville. Exposed there Lewis in 1954 found phytosaurian teeth and the plant is the following pertinent sequence of strata, omitting impressions described in the following pages, including the overlying 1000 feet of the Jurassic Wanakah and the object of unknown identity (pi. 33, fig. 4). Morrison formations, and 250 feet of the Cretaceous The nonmarine sediments of the Triassic in the "four Dakota sandstone: corners" area of the southwestern States are believed 364884—56 205 206 SHORTER CONTRIBUTIONS TO GENERAL ASSEMBLY

FIGURE 29. Tentative reconstruction of Sanmiguelia lewisi Brown. In life the leaves were probably somewhat stiff, like the lower right hand leaf, rather than undulant. No sheaths or crests are shown because discernible evidence for them is lacking. X 1/5 to have accumulated on floodplains and in scattered Numerous species of conifers have been described pools and lakes at not many hundreds of feet above sea from the Triassic of the United States, and particularly level. Such plants, therefore, as were entombed in the from the southwestern States, notably in and around sediments, most likely originated in the lowlands of the Petrified Forest National Monument near Hoi- the basin of deposition, whereas remains of more dis­ brook, Arizona. There is, therefore, nothing especially tant upland plants did not get into the fossil record. unusual about the presence of such remains in the Whatever the reason, whether land vegetation was Triassic of southwestern Colorado. sparse on account of climatic conditions, or whether the circumstances attending the accumulation of the MONOCOTYLEDONS sediments were unfavorable, plant remains in the SANMIGUELIA Brown, n. gen. Dolores apparently are relatively scarce. Plants with alternate, simple, large, elliptic, strongly DESCRIPTION OF THE SPECIMENS pleated, monocotyledonous, palmlike leaves. Stem CONIFERS rounded, tapering rapidly toward the apex, but other Brachyphyllum miinsteri Schenk features unknown. Plate 33, figures 1, 3 Sanmiguelia lewisi Brown, n. sp. Brachyphyllum miinsteri Schenk, 1867, Die fossile flora der Grenz- Plate 32, figures 1, 2; plate 33, figure 2 schichten des Keupers und Lias Frankens, p. 187, pi. 43, figs. 1-12. Elliptic to ovate entire, firm-textured leaves, attain­ The coniferous twigs from the Dolores formation, ing a length of 40 cm and width of 25 cm, conspicu­ here figured, resemble closely those shown by Schenk ously pleated, the strong ribs or veins converging from the Rhaetic formation at Bamberg, Germany. toward the apex and the base. Midrib not differen­ Neither these nor Schenk's specimens, however, quite tiated. Between the 16 to 20 ribs are closely spaced match the usual conception of the features of the genus parallel veinlets, as in all palms. The leaves are at­ Brachyphyllum with its closely appressed foliage. They tached to the stem alternately by broad, thin, clasping, can be compared with the twigs of some species of or probably sheathing petioles at vertical intervals of Palissya, Sequoia, Sphenolepidium, Araucarites, and 4 cm, as shown by three leaves preserved in one block others found in the Triassic and later formations. of sandstone. These permit a tentative reconstruction Without cones or other distinguishing features, the of the plant (fig. 29). However, neither sheaths nor identification of remains of this kind is fraught with crests (hastulas, ligules) are discernible, either because much uncertainty. Probably the specimen collected none existed or because the preservation around the by Cross (1899) in the Telluride quadrangle and identi­ stem is poor. The stem, now only a calcitic cast fied by David White as Pachyphyllum munsteri, was without cell structure, is rounded and in a vertical dis­ a representative of this species—the Pachyphyllum tance of 10 cm narrows rapidly from a diameter of 4 being an error for Brachyphyllum. Apparently, that cm to a blunt point, thus suggesting that it is the apical specimen is now lost. portion of a low, but mature plant.

PLATE 32 FIGURES 1, 2. Sanmiguelia lewisi Brown. Fig. 1 shows the broad base of the petiole clasping the grayish, granular, rounded stem. The latter is a calcitic cast without cellular structure. X % GEOLOGICAL SURVEY PROFESSIONAL PAPER 274 PLATE 32

TRIASSIC PLANTS FROM SOUTHWESTERN COLORADO GEOLOGICAL SURVEY PEOFESSIONAL PAPER 274 PLATE 33

TRIASSIC PLANTS FROM SOUTHWESTERN COLORADO PALMLIKE PLANTS FROM DOLORES FORMATION 207

The generic term Sanmiguelia is feminine and is from This is a further demonstration of the primitive nature of the the name of the river at the type locality. The species entire type * * *. In Latania commersoni J. F. Gmel. (L. borbonica Lam.), the seed-leaf is ovate, lanceolate (pi. 23, fig. is named for G. Edward Lewis, vertebrate paleontologist 94); during growth the later leaves become relatively broader, of the U. S. Geological Survey and discoverer of the until, by distal tension, splits take place (pi. 23, fig. 93), and in first specimens. the adult the leaf is a quite broad, fan-leaf type * * *. In the growth of a pinnate-leaved palm, when a new leaf first appears, RELATIONSHIPS OF SANMIGUELIA the pinnae cling together laterally on their outer margin and to / the distal leaflet, in such a way that the whole leaf simulates the Sanmiguelia leurisi is regarded tentatively but credi­ entire type and demonstrates the simple dynamic genesis of this bly as a primitive palm. Superficially its leaves re­ type of compound leaf from an ovate leaf like that of the young semble those of species of other monocotyledonous * * *. Also the distal tip of pinnate leaves repeats the form of genera, such as Carludovica, Curculigo, Cypripedium, nepionic leaves * * *. Some palms sucker from the base, and in such suckers we might expect to find localized stages in de­ Joinvillea, Lilium, Potamogeton, Setaria, Smilacina, velopment, as is the case. In Raphis flabelliformis L'Herit., a Veratrum, and others. In some of these the plicate or fan-leaf palm, suckers from the base have simple leaves like pi. accordion-pleated feature of the foliage is almost as 23, fig. 94, and only in later growth are the typical fan-leaves of strongly and comparably developed, and there may be the species acquired * * *. Suckers of Chrysalidocarpus no distinct midrib. Consequently, it is regrettable that lutescens repeat the form of the seedling so closely that they are no flowers, fruits, or stems with cell structure, have practically identical in leaf characters * * *. been found with Sanmiguelia to eliminate the foregoing To Jackson's discussion of juvenile leaf forms should genera completely and to establish its identity as a now be added Eames' recent interesting observations palm conclusively. on the development of leaves in the mature stages of Besides the suggestive morphologic appearance of palms (1953). In general, the simple seedling leaves its leaves, further reason for believing that Sanmiguelia are succeeded, according to the kind of palm, by pin­ represents a palm is derived from consideration of the nate or palmate leaves. Study of histological prepar­ probable ancestral features of primitive palms. The ations of leaf primordia indicates that the palmate is late Robert T. Jackson, from observations of the derived from the pinnate form, the central pinna of the juvenile leaves of palm seedlings and with numerous palmate being the distal pinna of the pinnate leaf, and lucid illustrations, discussed this phase of the subject the rachis of the palmate leaf being shortened or tele­ in part, as follows (Jackson, 1899, p. 124, pi. 23): scoped. The rein or marginal flange that is present Amongst palms many species present interesting features of on erupting leaves of mature plants is seldom seen on localized stages in development. I have studied the develop­ the leaves of juvenile plants. ment of seedling palms of thirty-seven species and from published Ancestral or reminiscent features have not been as figures of von Martius and others know something of the early definitely recognized in the lineage of the palms as they characters of forty-five species distributed through twenty-eight have in such genera as Ginkgo (Brown, 1939, p. 246; genera. Seedlings of all species have simple, early seed-leaves, and these are always one of two types. The first and most 1943, p. 862, figs, la-lc) and Cercidiphyllum (Brown, primitive type of seed-leaf is elongate, lanceolate, or oval, ter­ 1939). Until now the fossil record of the palms has minating more or less acutely, and with a longer or shorter dated from the Jurassic Lias. Lignier (1895, 1908), rachis, usually long. This type is seen in Latania (pi. 23, fig. from collections in Normandy, described and illustrat­ 94), Corypha (pi. 23, fig. 104), Cocos (pi. 23, fig. 102), Phoenix ed as Propalmophyllum liasinum, fragmentary portions (pi. 23, figs. 100, 103, 103a), and many other genera. The sec­ ond type of nepionic or seed-leaf is similar to the first, except of fan-shaped leaves showing chiefly the apexes of the that it is distally cleft. It is characteristic of Chrysalidocarpus petioles and bases of the fans, with a few ribs. Of (pi. 23, fig. 95), Kentia (pi. 23, fig. 96), Caryota (pi. 23, fig. 98), these Seward (1931, p. 366) said- and many other genera. Very young specimens of genera which Some French fossils described by Prof. Lignier as Propalmophyl­ have distally cleft leaves or exceptional individuals show a trans­ lum bear a striking resemblance to pieces of a palm leaf, though ient distal fusion of the apices of the leaf, as in Kentia (pi. 23, they cannot be accepted as proof of the existence of palms in a fig. 96), demonstrating that the cleft type is only a modification Jurassic flora. of the first or entire type. In some types as Stevensonia grandi- folia J. Dune. (Phoenicophorum sechellarum H. Wendl.), which Nevertheless, in a table on page 414 of the same book, have distally cleft leaves, the first leaves are entire, the cleft Seward indicates Palmae as dating from the Jurassic. character appearing later (third leaf of the specimen in hand). I, myself, see no reason for doubting Lignier's identi-

PLATE 33

FIGUKES 1, 3. Coniferous twigs of Brachyphyllum munsteri Schenk. X 1 2. Sanmiguelia lewisi Brown, showing the numerous parallel veinlets between the prominent ribs. X % 4. Unknown object, probably an animal trail. X 1 5. Paloreodoxites plicatus (Lesquereux) Knowlton, from the Paleocene Dawson arkose near Ramah, Colo. X 2 208 SHORTER CONTRIBUTIONS TO GENERAL ASSEMBLY fication of his leaves as palms. True, some fragmen­ Whatever the answer to the foregoing question, the tary Jurassic and later fossils, superficially somewhat leaves of Sanmiguelia resemble somewhat those of resembling the fragments figured by Lignier, have been Paloreodoxites and some from the Stonesfield slates in mistaken for remains of palms; but these are the egg the Jurassic Oolite at Eyeford, England, called Lilia capsules of chimaeroid fishes (Brown, 1946). [an error for Lilium] lanceotata Buckman (in Murchison, The first authentic remains of palms found in the .1845, p. 93, pi. 2, fig. 3). This is Buckman's descrip­ United States are from the early Upper Cretaceous. tion— From then on, both fan and feather palms are not Leaf ovato-lanceolate, with a slight acuminated point—leaf• uncommon. The fossil wood of palms, known as stalk short. An elegant parallel veined endogenous leaf, having Palmoxylon, is fairly abundant at some localities. much the appearance of leaves of the liliaceous tribe of plants. When thoroughly silicified it is attractive and durable; On the same plate Buckman figured two somewhat and in Aboriginal American days it was sometimes similar but more fragmentary leaves as Naiadea used for making arrow-points and other artifacts. acuminata and Stricklandia acuminata, both of which Fossil trunks, or portions thereof, that are sometimes he described as endogenous leaves. Seward (1904, misidentified as Palmoxylon, are the Mesozoic ferns p. 121, 122) synonymized "Lilia" and Naiadea with known as Tempskya. The fossil fruits of palms, par­ Podozamites stonesfieldensis and Stricklandia with Baiera ticularly of Nipa, are known. Others, sometimes phillipsi. As I have not seen these specimens I can called Palmocarpon, are somewhat indefinite and form an independent opinion of them only from doubtful. Buckman's descriptions and illustrations. Considera­ More particularly pertinent to the discussion of the tion of these almost persuades me that perhaps Seward's relationships of Sanmiguelia lewisi are leaves of Paloreo- disposition of the specimens was not altogether appro­ doxites plicatus (Lesquereux) Knowlton (pi. 2, fig. 5) priate. Particularly, I am doubtful about Buckman's from Paleocene strata in the Denver basin, Colorado— pi. 2, fig. 3 of "Lilia". How Buckman, whose other the Denver formation at Golden; and the Dawson illustrations seem to be fairly accurate, could have arkose at Ramah. The small fragment from the Eaton drawn a coarsely ribbed leaf from a specimen that, if formation at Tercio, described by Knowlton (1917, p. Seward is correct, should have no ribs but only nu­ 287, pi. 63, fig. 1), must be regarded as doubtfully merous closely spaced parallel veins, as in typical identified. These elliptic to ovate leaves (approxi­ Podozamites, I do not comprehend. His illustration mately 20 cm in length and 7 cm in width, with 12 looks too much like those of small leaves of Paloreo­ ribs) were called Oreodoxites plicatus by Lesquereux doxites, Sanmiguelia, or seedling leaves of living palms, (1883, p. 122, pi. 18, figs. 1-4). Knowlton (1930, p. 41, to be dismissed too lightly as Podozamites. Moreover, pi. 11, figs. 1-4) changed the generic name of this Seward (1904, p. 152, pi. 11, figs. 5, 6) described and species to Paloreodoxites because the genus Oreodoxites figured as Phyllites a poplarlike leaf from the same was founded by Goppert on an indefinite Permian seed, beds. This leaf has the appearance of an authentic and because Lesquereux may not have known of dicotyledonous leaf, and if it is such, it strongly sup­ Goppert's genus but thought he was creating the name ports the opinion that primitive palms, or at least anew on the basis of similarities he saw between his monocotyledons, also could have existed simulta­ leaves and those of the living Oreodoxa. Actually, as neously somewhere and perhaps in the same flora.1 there appears to be no close relationship between Furthermore, the fan palms (Propalmophyllum) of Paloreodoxites and Oreodoxa, Knowlton's name is not Lignier were already in existence in Lias time in too happy a substitute. Normandy. The doubtful nature of "Lilia," and the The prime question now is, Are the leaves of Paloreo­ impossibility of demonstrating any but a superficial doxites plicatus leaflets of a large compound leaf, as resemblance among "Lilia," Paloreodoxites, and the Lesquereux seems to have conjectured; simple, mature specimens here being reported, necessitate a new name leaves; or simple, seedling leaves? Because none of for the Colorado monocotyledon. these leaves has been found attached to a rachis or Possibly Sanmiguelia lewisi represents some plant stem, their exact nature is open to speculation. The other than a palm or similar monocotyledon with leaves could be simple seedling leaves of fan or of parallel-veined leaves. For example, Cordaites has feather palms, both kinds having been present in the strap-shaped leaves whose veins are closely spaced but Denver basin and adjoining areas during Paleocene the leaves are not plicate or strongly ribbed. Perhaps time. The leaves could be mature leaflets of a pinnate the yuccalike leaves sometimes reported from Triassic palm, or they could be mature leaves of a primitive i Under the name Sassendorfites benkerti, Oskar Kuhn, in a report just received from Germany, describes and illustrates another dicotyledonous leaf, this from lower kind of palm, a leftover from early Mesozoic time. Jurassic strata at Sassendorf, near Bamberg. (Orion, Jahrg. 10, no. 19-20, Oct. 1955.) PALMLIKE PLANTS FROM DOLORES FORMATION 209 localities under the name Yuccites are cordaitean origin of the angiospermous plants. Adequate sum­ foliage. Some cycads have leaves comparable in size maries of the speculations on this matter may be found to those of Sanmiguelia but the veins are not ribbed as in Arnold (1947, p. 333-365) and Axelrod (1952), with in palms, and, if closely parallel, may fork or anasto­ pertinent bibliographies. Sanmiguelia lewisi does not mose to make a notably reticulate pattern. So far as seem to provide answers to some of the fundamental I am aware, no other extinct or living plants, except questions involved in this speculation: Did the mono­ palms and species of the monocotyledons already cotyledons come from the dicotyledons or the dicotyle­ mentioned, have leaves that match or closely resemble dons from the monocotyledons or did both descend those of Sanmiguelia lewisi. This species, if not a from a common ancestor? Did woody monocotyle­ primitive palm, is a palmlike monocotyledon, whose dons derive from herbaceous ancestors? Did the simple leaves without midribs are fully developed, not angiosperms originate in upland areas, according to seedling leaves, of a mature plant. If this be true, Axelrod's thesis? As representatives of the angio­ these specimens are the oldest known megascopic sperms existed in the early Jurassic it is not altogether remains of the angiospermous flowering plants. surprising now to find their ancestors in the Triassic. Botanists and paleobotanists have long been looking Perhaps their ultimate origin must be sought in the for evidence concerning the time, place, and manner of Paleozoic.

BIBLIOGRAPHY

Arnold, C. A., 1947, An introduction to paleobotany: New York, Eames, Arthur, J., 1953, Neglected morphology of the palm McGraw-Hill Co. leaf: Phytomorphology, v. 3, no. 3, p. 172-189. Axelrod, D. I., 1952, A theory of angiosperm evolution: Evolu­ Hills, R. C., 1880, Note on the occurrence of fossils in the Triassic tion, v. 6, p. 29-60. and Jurassic beds near San Miguel in Colorado: Am. Jour. Brown, Roland W., 1939, Fossil plants from the Colgate member Sci., v. 19, p. 490. of the Fox Hills sandstone and adjacent strata: U. S. Geol. ——— 1882, Jura-Trias of southwestern Colorado: Am. Jour. Survey Prof. Paper 189, p. 239-275, pis. 48-63. Sci., v. 23, p. 243. ———— 1939, Fossil leaves, fruits, and seeds of Cercidiphyllum: Jackson, R. T., 1899, Localized stages in development in plants Jour. Paleontology, v. 13, p. 485-499. and animals, Mem. Boston Soc. Nat. Hist., v. 5, no. 4, ———— 1943, Some prehistoric trees of the United States: Jour. p. 89-153, pis. 16-25. Forestry, v. 41, p. 861-868, fig. 1. Knowltoh, F. H., 1917, Fossil floras of the Vermejo and Raton 1946, Fossil egg capsules of chimaeroid fishes: Jour. formations of Colorado and New Mexico: U. S. Geol. Paleontology, v. 20, p. 261-266, pis. 38, 39. Survey Prof. Paper 101, p. 223-455. ———— 1930, The flora of the Denver and associated formations Buckman, James, 1845, in R. I. Murchison, Outline of the of Colorado: U. S. Geol. Survey Prof. Paper 155. geology of the neighborhood of Cheltenham, p. 93-109, Lesquereux, Leo, 1883, The Cretaceous and Tertiary floras: pis. 1-13. U. S. Geol. Survey Terr. Rept., v. 8. Cross, Whitman, Howe. Ernest, and Irving, J. D., 1907, Ouray Lignier, Octave, 1895, Vegetaux fossiles de Normandie—II, Con­ quadrangle, Colorado: U. S. Geol. Survey, Geol. Atlas of tributions a la flore liasique de Sainte-Honorine-la-Guillaume the U. S., Folio 153. (Orne.): Mem. Soc. Linn. Normandie, v. 18, p. 121-152, Cross, Whitman, Howe, Ernest, and Ransome, F. L., 1905, figs. 1-6. Silverton quadrangle, Colorado: U. S. Geol. Survey, Geol. ———— Idem, 1908, V. Nouvelles recherches sur le Propal- Atlas of the U. S., Folio 120. mophyllum liasinum Lignier: Idem., v. 23, p. 1-14, pi. 1. Cross, Whitman, and Purington, C. W., 1899, Telluride quad­ Seward, A. C., 1904, Catalogue of the Mesozoic plants in the rangle, Colorado: U. S. Geol. Survey, Geol. Atlas of the Department of geology British Museum (Natural History). U. S., Folio 57. The Jurassic flora—II, Liassic and oolitic floras of England. Cross, Whitman, and Ransome, F. L., 1905, Rico quadrangle, ———— 1931, Plant life through the ages: New York, The Colorado: U. S. Geol. Survey Atlas of the U. S., Folio 130. Macmillan Co.

U. S. GOVERNMENT PRINTING OFFICE: 1956